Creative Claw

Generate on-brand media inside ChatGPT, Codex, Claude, Cursor, Grok Bot, Hermes, and OpenClaw.

Creative Claw is an MCP plugin that brings a full AI media studio into Grok Bot, Hermes Agent, OpenClaw, Cursor, Claude Code, Claude Desktop, ChatGPT, and Codex. Generate images, video, and expressive speech; reuse Characters and brand assets; and produce product campaigns, UGC ads, and multi-shot Films through one account. No API keys and no platform switching.

creativeclaw.co | Join the Beta |

What You Get

MCP Server

One connection to Creative Claw's MCP server gives the skills live model discovery plus media generation, editing, brand, asset, Character, Film, and feedback tools:

Category	Tools
Image	`generate_image` (generate, edit, or repeat across models for comparison), `load_image`
Video	`generate_video`, `check_job`
Speech	`generate_speech`
Media editing	`remove_background`, `upscale_media`, `trim_video`, `scale_video`, `add_subtitles`, `extract_frames`, `merge_media`
Models	`list_models`, `get_model_params`
Assets	`search_assets`, `update_asset`, `delete_asset`, `upload_asset`, `import_media`, `get_upload_url`, `confirm_upload`
Brand themes	`get_theme`, `list_themes`, `update_theme`, `delete_theme`
Characters	`manage_character`, `list_characters`, `clone_voice`, `delete_character`
Films	`create_film_project`, `update_film_project`, `get_film_project`, `list_film_projects`, `assemble_film`
Feedback	`submit_feedback`
Cost and account	`estimate_generation`; `get_credits_balance` and `get_credits_link` where the client exposes them

Access 1,000+ production-ready AI models — FLUX, Gemini, Veo, Sora, Kling, Seedance, Hailuo, HeyGen, Recraft, ElevenLabs, and more — through a single unified account with usage-based pricing.

Skills (Creative Workflows)

The creativeclaw root skill routes mixed or unclear requests. Focused outcome skills own general image, video, and voice work plus Characters, planning, Films, product photoshoots, UGC ads, and feedback. Model specialists add exact prompting and reference guidance only after an outcome chooses Nano Banana 2, Nano Banana Pro, GPT Image 2.5 Flare or Sunburst, Seedream 5 Pro, Gemini Omni, Seedance 2.5, H3 Max, ElevenLabs v3, or xAI TTS.

Every OpenAI skill declares the ChatGPT MCP dependency at https://app.creativeclaw.co/mcp/chatgpt, matching the endpoint configured in the plugin draft. A focused skill can therefore activate directly without losing access to Creative Claw's tools.

Supported Models

Recommended image models

Default to Gemini 3.1 Flash (Nano Banana 2) for most image generation and editing. It is the most cost-efficient recommendation overall, balancing quality, speed, and cost; use a specialist model only when its specific strength matters.

Model	Creative Claw ID	Best use
Nano Banana 2	`image/nano-banana-2`	Default and cost-efficient choice for most generation and editing
Nano Banana Pro	`image/nano-banana-pro`	Complex professional layouts, typography, and demanding composites
GPT Image 2.5 Flare	`image/gpt-image-2.5-flare`	Fast, high-quality everyday OpenAI image generation and editing
GPT Image 2.5 Sunburst	`image/gpt-image-2.5-sunburst`	Precision-focused generation, transparency, and tightly controlled edits
Seedream 5 Pro	`image/seedream-5-pro`	Premium product, fashion, and commercial imagery

Recommended speech

Use speech/elevenlabs-v3 by default for narration, dialogue, emotional delivery, and multilingual speech. Use speech/xai-tts when its 28 built-in voices, exact inline/wrapping performance tags, or G.711 telephony output are the better fit. Creative Claw also supports consent-gated ElevenLabs Instant Voice Cloning as a separate Character workflow.

Use list_models and get_model_params at runtime rather than assuming a fixed catalog or reference limit.

Install

Grok Bot and Cursor

Creative Claw is packaged for the Cursor Marketplace used by Grok Bot's plugin system. After marketplace approval, open Settings → Plugins → Marketplace in Grok Bot or Customize → Plugins in Cursor, search for Creative Claw, install it, and complete browser authentication on the first tool call.

For local Cursor testing before publication, symlink the plugin directory and reload Cursor:

ln -s /path/to/creative-claw-marketplace/plugins/creative-claw ~/.cursor/plugins/local/creative-claw

Hermes Agent

Hermes supports the Creative Claw hosted MCP directly. Run:

hermes mcp add creative-claw --url https://app.creativeclaw.co/mcp --auth oauth
hermes mcp login creative-claw

Or use the one-click Add to Hermes link on a machine with Hermes installed. The portable plugin.json and mcp.json in the plugin root also make the package compatible with Agent Plugins v1 clients. Until Hermes supports OAuth login for plugin-supplied MCP entries, the two hermes mcp commands above are the reliable installation path.

OpenClaw

Creative Claw includes a native OpenClaw manifest, the consolidated skill, and ClawHub package metadata. Once the package is published to ClawHub, install it with:

openclaw plugins install clawhub:@creativeclaw/plugin
openclaw plugins enable creative-claw
openclaw mcp login creative-claw
openclaw gateway restart

The first login opens Creative Claw's OAuth flow; no API key needs to be copied into a config file.

Agent skills (`npx skills`)

The canonical cross-client skill remains in the standard repository layout, so skill-directory users can install it directly:

npx skills add CreativeClawCo/creative-claw-marketplace

The OpenAI Store installs the shared ChatGPT/Codex plugin and skill bundle separately; Store users do not need this command.

Claude Code

# 1. Add the marketplace
claude plugin marketplace add CreativeClawCo/creative-claw-marketplace

# 2. Install the plugin
claude plugin install creative-claw@creative-claw-marketplace

# 3. Authenticate — on first use, the MCP server will prompt you to sign in via Clerk OAuth

That's it. The plugin connects to Creative Claw's MCP server and installs the consolidated creativeclaw workflow skill.

Claude Desktop

Add to your MCP config (claude_desktop_config.json):

{
  "mcpServers": {
    "creative-claw": {
      "type": "http",
      "url": "https://app.creativeclaw.co/mcp"
    }
  }
}

No API keys needed — auth is handled via Clerk OAuth on first connection.

Glama / Directory Introspection

The glama-mcp/ directory provides a lightweight stdio MCP wrapper for Glama.ai directory compatibility. This wrapper exposes 9 core tool descriptions for discovery (how_to_connect, generate_image, generate_video, generate_speech, list_models, get_model_params, check_job, list_characters, get_credits_balance) that guide users to connect to the full production MCP server at https://app.creativeclaw.co/mcp.

The Glama wrapper does not perform actual generation—it serves as a directory entry point. To use Creative Claw's full generation capabilities, connect to the production HTTP MCP server as described above.

Run locally:

cd glama-mcp && npm install && npm start

Build Docker image:

docker build -t creative-claw-glama .
docker run -i creative-claw-glama

Quick Start

Talk to the host naturally:

"Generate a product photo of my headphones on a marble surface, golden hour lighting"
  -> creativeclaw routes to its image workflow, picks the model, and generates

"Make a 15-second cinematic video of coffee being poured in slow motion"
  -> creativeclaw routes to its video workflow, generates a reference image, then the video

"Edit this image — change the background to a beach sunset, keep the person unchanged"
  -> creativeclaw routes to an edit model, preserves identity, swaps background

"Set up our brand — here's our website"
  -> creativeclaw extracts colors/fonts/logos, uploads the assets, and saves a reusable theme

"I need a TikTok-style product video for this shoe" [attach image]
  -> the UGC workflow plans the ad, creates reference frames, selects video and voice models, and assembles a first cut

"Create a reusable presenter from this portrait and clone my voice; I confirm it is mine"
  -> the Character and consent-gated ElevenLabs voice-cloning workflows create and audition the reusable identity

"The video tool ignored my end frame—please report it"
  -> the feedback workflow submits one actionable bug report with the attempted task and tool

Architecture

You -> Host + outcome skill (goal, approvals, creative direction)
              |
      Model specialist (prompt and reference details when selected)
              |
         MCP Tools (generation, assets, Characters, Films, feedback)
              |
       Creative Claw Server (model providers + R2 storage + Clerk auth)
              |
       Permanent media URLs (never expire)

No API keys. No CLI wrappers. No expiring URLs. Just skills + MCP.

Project Structure

This repo is a marketplace — it contains one or more installable plugins.

.claude-plugin/
  marketplace.json         # Marketplace manifest (required for Claude Code marketplace sync)
  plugin.json              # Root plugin metadata
.cursor-plugin/
  marketplace.json         # Cursor/Grok Bot marketplace manifest
plugins/
  creative-claw/
    .cursor-plugin/
      plugin.json          # Cursor/Grok Bot plugin manifest
    .claude-plugin/
      plugin.json          # Plugin manifest (MCP server config)
    .mcp.json              # Grok Build/Codex MCP server config
    plugin.json            # Portable Agent Plugins v1 manifest (Hermes-compatible)
    mcp.json               # Portable Agent Plugins v1 MCP config
    openclaw.plugin.json   # Native OpenClaw manifest
    skills/
      creativeclaw/
        SKILL.md           # thin router for mixed and workspace requests
        references/        # shared asset, theme, upload, and editing workflows
      creativeclaw-*/      # outcome and recommended-model specialist skills
skill-variants/
  chatgpt/
    platform-upload.md     # OpenAI Store routing for ChatGPT attachments and Codex local files
scripts/
  build-skill-zips.sh      # builds both upload-ready skill archives
creativeclaw-skill.zip             # cross-client archive
creativeclaw-chatgpt-skill.zip     # OpenAI Store root-skill archive
creativeclaw-*-chatgpt-skill.zip   # focused OpenAI skill archives
evals/skill-routing-scenarios.md   # activation and workflow regression suite

Maintaining the two distributions

Edit the canonical skill once under plugins/creative-claw/skills/creativeclaw. Put only ChatGPT-specific routing differences in skill-variants/chatgpt/platform-upload.md, then run:

./scripts/build-skill-zips.sh

The build validates every skill's metadata and MCP dependency, checks routing coverage and the regression suite, and creates deterministic upload archives. Nothing in the build publishes or submits a plugin draft.

Pricing

Usage-based — pay only for what you generate. No subscriptions, no commitments. Check creativeclaw.co for current rates.

Compatibility

Grok Bot — via the Cursor marketplace plugin and hosted MCP server
Cursor — via .cursor-plugin/plugin.json
Claude Code — via .claude-plugin/plugin.json
Claude Desktop — via MCP server config
Codex and other skill-directory clients — via the canonical creativeclaw skill
OpenAI Store (ChatGPT + Codex) — via creativeclaw-chatgpt-skill.zip
Hermes Agent — via OAuth MCP setup and portable Agent Plugins v1 manifests
OpenClaw — via openclaw.plugin.json and ClawHub-ready package metadata

All use the same skills and connect to the same MCP server.

License

Apache-2.0

Built by Creative Claw Co.

Read https://aiagentslisting.com/mcp/creative-claw to learn what the "Creative Claw" MCP server does and how to install it. Add it to my coding agent's MCP configuration as documented on that page, then confirm the server connects and list the tools it exposes.

Name

What it does

mcp_ui_action

Route grouped Creative Claw UI-only actions. notify_job requests one completion email after the user clicks Notify me. track_ui_event accepts narrowly scoped UI event payloads, but optional UI analyti

generate_image

Generate or edit images using AI models. Use this for AI-generated visual assets, including branded social cards, banners, posters, product images, and images guided by a saved theme or reference imag

compare_models

Generate the same image with multiple models side-by-side for comparison. All models receive the same prompt and settings, and results are displayed together in a single view. Use this when the user

generate_video

Costly operation: submit an AI video generation using the user's Creative Claw credits. Requires explicit user approval before every generation. Returns a completed video or a job ID for background pr

generate_speech

Generate speech from text with ElevenLabs, Cartesia, or another listed speech model. Returns completed media or a queued job ID that resolves through check_job. ElevenLabs Multilingual v2 (model: "sp

generate_audio

Generate a sound effect, ambience, or music track with ElevenLabs. Returns a permanent audio URL with an inline player. - Sound effects and ambience: model sfx/elevenlabs-sound-v2. Prompts are limite

transcribe

Transcribe audio or video to text with ElevenLabs Scribe. Direct Scribe supports hosted audio/video, YouTube, TikTok, Instagram, and other public video-hosting URLs when the provider can fetch them. I

isolate_audio

Clean up an audio file using ElevenLabs Voice Isolator — removes background noise, music, and reverb so only the voice remains. **Workflow:** in ChatGPT, call `import_chatgpt_media` for an audio file

render_html_image

Render HTML/CSS to a PNG image via headless Chromium. Use this when the user explicitly asks to render HTML/CSS, provides HTML, or requests a deterministic, pixel-controlled layout assembled with web

render_html_video

Render an HTML/CSS/JS composition to an MP4 video using HyperFrames on Modal.com. Use only when the user explicitly asks for HTML-to-video, HyperFrames, code-driven motion, supplies animated HTML, or

render_template

Render a saved template one or more times with different parameter values. Returns PNG images. **Multi-variant rendering (the common case):** pass `renders` as an array — each entry is the per-varian

remove_background

Remove the background from an image or video using AI. Returns a permanent URL to the result. - For **images**: produces a transparent PNG. Just provide the URL and type=image. - For **videos**: uses

upscale_media

Upscale an image or video to higher resolution using AI. Returns a permanent URL to the upscaled result. **Images** — three models: - **aura** (default): Cheapest and fastest, fixed 4x upscale. Just

trim_video

Cut one continuous time range from either a video or an audio file. Pass exactly one of `video_url` or `audio_url`. Existing video trimming remains backward-compatible and returns the completed video.

cut_and_reframe_video

Create one edited MP4 by cutting and reordering selected timestamp ranges from a workspace video, preserving original audio, and reframing each cut for portrait, landscape, square, or custom output di

scale_video

Queue a video resize and return a job ID immediately. This is a standard FFmpeg resize operation—not AI upscaling—and does not add visual detail. Call check_job with the job ID for the permanent scale

add_subtitles

Auto-transcribe and burn karaoke-style subtitles onto a video. Returns a permanent URL to the subtitled video. Features word-level highlighting (karaoke effect), compatible Google Font overrides, cus

extract_frames

Queue frame extraction from a video and return a job ID immediately. Call check_job with the job ID for the permanent extracted-frame URLs. Modes: - **single**: Extract one frame: first, middle, or l

merge_media

Queue a media merge and return a job ID immediately. The merge runs in the background; call check_job with the returned job ID when you need the permanent output URL. Operations: - **merge_audio_vide

load_image

Load an image from a URL and return it as base64 so you can see it in your context. Use this ONLY when: - The user explicitly asks you to look at / review an image - You need to iterate on a generated

list_models

List available AI models, filtered by category or search query. Categories: - "image" — models that generate and/or edit images - "video" — models that generate video from text and/or images - "speec

get_model_params

Get all available input parameters for a specific AI model. Returns the full schema including parameter names, types, defaults, constraints, and descriptions. Use this to discover model-specific para

estimate_generation

Estimate the credit cost of one generation before submitting it, and compare that estimate with the current user's balance. This tool is read-only: it does not generate media or deduct credits. Use i

search_examples

Search Creative Claw's curated prompt examples for inspiration or a close starting point. Use when the user asks for examples, references, prompt ideas, a particular creative style, or something simil

get_example

Retrieve the complete prompt or executable HyperFrames source for one Creative Claw example selected from search_examples. For renderType html_video, sourceType is html or zip: renderSource contains e

check_job

Check the status of a generation job. Returns the current status and, when completed, the permanent media URL. Call this after generate_image, generate_video, or any asynchronous media utility to pol

import_media

Open an interactive upload UI for the user to choose images, videos, audio, fonts, or ZIP archives from their device. Call this tool when: - The user has local files and wants to choose them through

search_assets

Search your media library. Assets include images, videos, audio, 3D models, documents, and ZIP archives that were generated, uploaded, or imported — each has a permanent URL, optional name, tags, and

update_asset

Organize a media asset by setting its name, tags, or description. Assets are entries in your media library (images, videos, audio, 3D models). - **Name**: Unique per user — lets you reference assets

delete_asset

Permanently remove an asset's stored file from Creative Claw storage and hide its library record. The asset will no longer appear in search results and its name is freed for reuse. This cannot be undo

upload_asset

Add a file to your asset library by direct URL. This tool supports images, videos, audio, 3D models, and ZIP archives. The type is inferred from content_type when omitted and the MIME family is unambi

get_upload_url

Get a presigned URL for uploading local file bytes directly to storage. Use this when the client can read a local file and make the PUT request itself, especially Codex and other local execution envir

confirm_upload

Confirm that a file has been uploaded via the presigned URL from get_upload_url. Verifies the file exists in storage and activates the asset so it appears in search results. Call this only after the

get_theme

Fetch a brand theme. Themes are reusable brand configuration bundles (colors, fonts, logos, product images, etc.) stored as JSON — use them to keep generated media on-brand. Returns the theme's name,

update_theme

Create or update a brand theme. Themes store colors, fonts, logos, a markdown `notes` field (voice / audience / tone / do's-and-don'ts), a reference image, and any brand elements used for consistent m

list_themes

List all brand themes. Returns each theme's name, default status, and a summary of stored keys (colors, fonts, logos, etc.). The MCP client renders results as a thumbnail carousel — click a card to c

delete_theme

Delete a brand theme (soft-delete). If the deleted theme was the default, the oldest remaining theme is promoted.

manage_character

Create or update a Character — a reusable persona with a description and reference image. **To create a character:** 1. Collect name + description (appearance, personality, role) 2. Generate a refere

clone_voice

Create a provider-hosted Instant Voice Clone from a private audio_asset_id or a publicly downloadable audio_url. Attach it to character_id when supplied; otherwise omit character_id and optionally pas

list_characters

List all Characters (reusable personas). Returns each character's id, title, description, whether it has a cloned voice, and its reference image. Pass a character's id as character_id to generate_imag

delete_character

Delete a Character, revoke voice access, and delete its current and retired provider voice clones and private source samples. Generated speech remains a separate asset. Retains minimal consent/audit r

create_film_project

Start a character-driven film project. Creates the project shell (status "drafting") and opens the film preview. Next: draft a script + shot list and save it with update_film_project, then show it to

update_film_project

Update a film project: save the script (logline + shots), advance status, attach generated assets, or change the cast/theme. Shots: pass the full `shots` array to replace the shot list (e.g. when fir

get_film_project

Load a film project and show its preview (cast, script, shots, assembled cut).

list_film_projects

List your film projects (newest first). Click a card to copy its id.

assemble_film

Queue an assembled first cut for an existing Film project by concatenating every rendered shot clip in order, optionally overlaying the project's single audioUrl narration track. Mode "connect" is the

create_template

Create a reusable template. Two kinds: - **html** — HTML/CSS with {{token}} placeholders rendered via headless Chromium → PNG. Provide html, width, height, parameters. Load fonts directly in the temp

update_template

Update an existing template's fields. Pass template_id or template_name, then any subset of fields to overwrite. Fields you omit are left alone. Kind cannot be changed after creation. **Replacement s

list_templates

List the caller's reusable image templates. Use this to discover templates by name before calling render_template, e.g. "generate a LinkedIn card announcing X" → list_templates → find the LinkedIn one

submit_feedback

Report product feedback about Creative Claw — bugs, missing capabilities, confusing flows, or praise. Use this when the user asks to report feedback. You may also suggest it when you observe meaningf

Name

What it does

Examples Browser

Search, preview, and use Creative Claw prompt examples.

Media Viewer

Interactive viewer for generated images, videos, and audio

Media Viewer (legacy)

Interactive viewer for generated images, videos, and audio

List Viewer

Carousel of templates or themes — click to copy the name (or ID).

Theme Editor

Visual theme editor — drop fonts, colors, logos, and a reference image; click save.

Theme Editor (legacy)

Visual theme editor — drop fonts, colors, logos, and a reference image; click save.

Character Editor

Visual character editor — set title/description/image and upload or record a voice to clone; click save.

Film Preview

Film project preview — assembled cut, cast, narration, storyboard/shot strip, and approval-gate actions.

Upload Media

Interactive drag-and-drop uploader for images and videos

Connect to the AI Agents Listing MCP server at https://aiagentslisting.com/api/mcp using the HTTP transport and complete the sign-in when it asks. Then call the submit_review tool for the listing "creative-claw" with a rating from 1 to 5 and a short note on what worked and what did not.

Model	Creative Claw ID	Best use
Gemini Omni	`video/gemini-omni-flash`	Default general video, native audio, references, and source-video work
Seedance 2.5	`video/seedance-2.5`	Premium, longer, reference-rich cinematic work
Seedance Mini	Discover with `list_models`	Economical drafts and iteration
MiniMax H3 Max	`video/minimax-h3-max`	Fast cinematic native-audio clips and multimodal references
H3 Max Fast	`video/minimax-h3-max-turbo`	Faster, lighter H3 Max iteration

Creative Claw

Creative Claw

What You Get

MCP Server

Skills (Creative Workflows)

Supported Models

Recommended image models

Recommended video models

Recommended speech

Install

Grok Bot and Cursor

Hermes Agent

OpenClaw

Agent skills (`npx skills`)

Claude Code

Claude Desktop

Glama / Directory Introspection

Quick Start

Architecture

Project Structure

Maintaining the two distributions

Pricing

Compatibility

License

Install with your agent

From the project README

Tools50

Resources9

What's missing

Show this grade in your README

Alternatives to Creative Claw

Send this prompt to your agent to leave a review

Subscribe to our newsletter

Latest articles

Creative Claw

Creative Claw

What You Get

MCP Server

Skills (Creative Workflows)

Supported Models

Recommended image models

Recommended video models

Recommended speech

Install

Grok Bot and Cursor

Hermes Agent

OpenClaw

Agent skills (npx skills)

Claude Code

Claude Desktop

Glama / Directory Introspection

Quick Start

Architecture

Project Structure

Maintaining the two distributions

Pricing

Compatibility

License

Install with your agent

From the project README

Tools50

Resources9

What's missing

Show this grade in your README

Alternatives to Creative Claw

Send this prompt to your agent to leave a review

Subscribe to our newsletter

Latest articles

Agent skills (`npx skills`)